The nature and use of proto-languages

Abstract

While it is true that languages related through (as one says) common descent are derived from a common ancestor language, this idiom, if unattested and available only in the form of a reconstructed proto-language, has only a limited degree of realism. The manner in which linguistic reconstruction proceeds, the lack of truly quantitative criteria in determining type and degree of linguistic relatedness, and the necessity to arrive at an entirely uniform linguistic construct are all apt to deliver a distorted or false view of the proto-language. Indeed its very existence may be questionable, especially if it cannot be supported by non-linguistic evidence; this applies in particular to intermediate reconstructed proto-languages like Proto-Italo-Keltic, Proto-West-Germanic, Proto-Ibero-Romanic, which are but way-stations on the road to the ultimate parent language. It is therefore suggested that all proto-languages be considered creations for the convenience of linguistic investigation and for the purpose of delving into an otherwise inaccessible linguistic past, but that no claim should be made for their being viewed and dealt with as real languages in any sense of the word, unless and until there accrues sufficient non-linguistic evidence for fixing them in time and place and for associating them with an anthropologically, archaeologically, or historically identifiable society. The reverse process, that is, the creation of a society to go with an unattested, reconstructed proto-language, is altogether improper.